基于特征元素的圖像檢索方法
Image Retrieval Based on Feature Element
-
摘要: 該文提出了一種基于特征元素的新的圖像儉索算法。持征元素與特征向量相比更注重根據(jù)人的主觀感知來表達(dá)圖像的視覺特征.在特征元素的基礎(chǔ)上,該文先定義了圖像間的相似性度量,即特征元素間的距離,又分別實(shí)現(xiàn)了對(duì)不同特征元素類別的距離計(jì)算。檢索實(shí)驗(yàn)表明,基于特征元素的圖像檢索算法能夠取得更符合人們視覺感知的結(jié)果。
-
關(guān)鍵詞:
- 圖像檢索; 特征元素; 特征向量; 圖像相似度
Abstract: In this paper, a novel image retrieval algorithm based on feature element is proposed. Compared with feature vectors, feature elements tend to represent visual meanings of the image according to subjective perception of human beings. Based on feature elements, the similarity between images is defined which is measured by the distant between feature elements. Retrieval experiments show that image retrieval based on feature elements can retrieve images which are more accordant with visual perception of human beings. -
Mukherjea S, Cho J H. Automaticallydetermining semantics for World Wide Web multimedia information retrieval[J].Journal of Visual Languages and Computing.1999, 10(6):585-606[2]Smith J R, Chang S-F. Visually searching the Web for content[J].IEEE Multimedia Magazine.1997, 4(3):12-20[3]Xu Y, Zhang Y-J. Feature element theory for image recognition and retrieval[C]. Proc. Storage and Retrieval for Media Databases, SPIE, 2002, Vol.4676: 126-137.[4]Xu Y, Zhang Y-J. Image retrieval framework driven by association feedback with feature elements evaluation built in[C]. Proc. Storage and Retrieval for Media Databases, SPIE, 2001, Vol.4315:118-129.[5]李勍,章毓晉.基于特征元素的圖像分類方法.電子學(xué)報(bào),2002,30(9):1262-1265.[6]Yao Y-R.[J].Zhang Y-J. Shape-based image retrieval using wavelets and moments[C]. Proc. of Workshop on Very Low Bitrate Video99, Kyoto, Japan.1999,:-[7]Pass G, Zabih R. Histogram refinement for content-base image retrieval[C]. Proc. of the IEEE Workshop on Applications of Computer Vision, Sarasota, FL, USA, 1996: 96-102.[8]Huang J, et al.. Spatial color indexing and applications[C]. Proc. of the IEEE International Conference on Computer Vision, Princetor, NJ, USA, 1998: 602-607. -
計(jì)量
- 文章訪問數(shù): 2177
- HTML全文瀏覽量: 87
- PDF下載量: 716
- 被引次數(shù): 0
下載:
